Learning Center
Plans & pricing Sign in
Sign Out

Exercise 3 OPenDAP and LAS technology applied to - MEOM TEAM


  • pg 1
									                        Exercise 3
    OPenDAP and LAS technology applied to Hurricane Isabel

The aim of this exercise is twofold. First, it will give you an overview of the available tools
used to access operational forecast systems outputs. They are the Live Access Server (LAS)
and the OPeNDAP (DODS) servers. Second, you will use those tools to examine the ocean
response to the hurricane Isabel. The MERSEA project gathers several operational systems
for the North Atlantic and is using the OPeNDAP and LAS technologies. You will use this
framework to examine the ocean response to the Isabel hurricane which traveled from north
of the Bahamas to North Carolina (US) in September 2003.

We first give you an introduction to the Live Access Server and OPeNDAP Server, and then
present the MERSEA project. Presently, ocean data products can be accessed using either a
Live Access Server or via an OPeNDAP request. The MERSEA and GODAE projects are
using the DMAC/IOOS OPeNDAP technology. The Data Management and Communications
committee (DMAC) of the Integrated Ocean Observing System (IOOS) has put together a
data management plan. The OPeNDAP data access protocol is at the core of the DMAC/IOOS
in order to present coherent data management strategy that can effectively integrate data
streams across disciplines and spatial and temporal scales. DMAC within the marine
environment presents great challenges due to the variety and complexity of the observations
that are involved. The resulting benefits are important, such as improved climate forecasts and
more effective protection of coastal marine ecosystems.

Live Access Server
The Live Access Server (LAS) is a highly configurable Web server designed to provide
flexible access to geo-referenced scientific data. It can present distributed data sets as a
unified virtual data base through the use of OPeNDAP/DODS networking (see next section).
Ferret is the default visualization application used by LAS, though other applications (Matlab,
IDL, GrADS, ...) can also be used.
LAS enables the Web user to:
     visualize data with on-the-fly graphics
     request custom subsets of variables in a choice of file formats
     access background reference material about the data (metadata)
     compare (difference) variables from distributed locations
LAS enables the data provider to:
     unify access to multiple types of data in a single interface
     create thematic data servers from distributed data sources
     offer derived products on the fly
     remedy metadata inadequacies (poorly self-describing data)
     offer unique products (e.g. visualization styles specialized for the data).

More information about LAS can be found on the web:


OPeNDAP has developed a software framework that simplifies all aspects of scientific data
networking, allowing simple access to remote data. Local data can be made accessible to
remote locations regardless of local storage format by using servers. Existing, familiar data
analysis and visualization applications can be transformed into clients (i.e., applications able
to access remote served data).
Although the OPeNDAP software was originally designed and developed by oceanographers
and computer scientists for oceanographic data (as part of the DODS, the Distributed
Oceanographic Data System), there is nothing in the design of the OPeNDAP software that
constrains its use to oceanography. Indeed, it is being considered by segments of the
meteorological and space science communities.

OPeNDAP allows you to access remote data over the internet (you are an OPeNDAP client):
    using familiar data analysis and visualization packages like: Matlab, Ferret, IDL and
    without worrying about data storage formats
    accessing only the data of interest

OPeNDAP allows you also to make your data available remotely (you are an OPeNDAP
server), regardless of the format of your data.
In this exercise, you will only use OPeNDAP as a client by accessing remote databases.
More information about OPeNDAP/DODS can be found on the web :
An non exhaustive list of all datasets available via OPeNDAP servers is indicated on the
OPeNDAP homepage:

What difference between LAS and OPeNDAP?
A LAS is an interface sending requests to OPeNDAP servers in order to provide to the user:
     data files
     plots (1D or 2D)
     It is also possible to make comparisons between 2 datasets and to perform simple
        mathematical operations (like average, difference, variance calculations…)
An OPeNDAP server is designed to provide data to any client sending requests. This client
may be a LAS or a user accessing the OPeNDAP server via requests sent by visualization
software like Matlab, GrADS, IDL or Ferret. In this exercise you will use Ferret to access
datasets hosted by different OPeNDAP servers.

The Ferret software
To access OPeNDAP servers, you will use the Ferret data visualization and analysis software.
Ferret is an interactive computer visualization and analysis environment designed to meet the
needs of oceanographers and meteorologists analysing large and complex gridded data sets. It
runs on most Unix/Linux systems, and on Windows NT/9x using X windows for display. It
can transparently access extensive remote Internet data bases using OPeNDAP (formerly
known as DODS);

Ferret was developed to analyse the outputs of numerical ocean models and compare them
with gridded, observational data. The model data sets are generally multi- gigabyte in size
with mixed 3 and 4-dimensional variables defined on staggered grids. Ferret offers a

Mathematica-like approach to analysis; new variables may be defined interactively as
mathematical expressions involving data set variables. Calculations may be applied over
arbitrarily shaped regions. Fully documented graphics are produced with a single command.

Many excellent software packages have been developed recently for scientific visualization.
The features that make Ferret distinctive among these packages are Mathematica-like
flexibility, geophysical formatting, "intelligent" connection to its data base, memory
management for very large calculations, and symmetrical processing in 4 dimensions. Further
information can be found on the web:

Ferret in a few words:
Here is a general overview of ferret.

Ferret used as visualization tool:
Ferret is designed to perform data visualization and data analysis. In this exercise, you will
mostly use the graphic capacities of Ferret. First, all datasets that you will use are Netcdf files
containing one or several variables. Each variable (like a temperature or a salinity field) is
handled as a multi-dimensional array, with generally 4 coordinates: X, Y, and Z for the space
and T for the time. If you want to make a 2D graphic with a 4D variable you have to select 2
axes out of 4 and perhaps to specify a sub domain (to make a zoom) before making the plot.

The OPeNDAP technology gives the opportunity to access remote Internet data bases
transparently from Matlab, GrADS, IDL or Ferret. Contrary to the other softwares, Ferret
does not require the installation of a „DODS Toolkit‟ to access OPeNDAPs as it integrates by
default this function.

Here are some commands that you will have to use in the exercise.
Consider the Levitus climatology stored on the MERCATOR OPeNDAP Server, and let‟s
make a plot of the salinity field over the North Atlantic at 5m depth.
 1) Launch Ferret by simply typing “ferret”
 2) Let‟s load the annual mean Levitus climatology

use “”

 Let‟s have a look at the available variables:
 show data
 You can see that this file contains 2 variables (Temperature and Salinity) and that each variable
 consist of a 3D array (869 points in longitude -i index-, 441 points in latitude -j index- and 12
 vertical levels -k index-). As it is a mean climatology, it does not have any temporal axis (L index
 for the time)
 3) Let‟s make a first plot of the North Atlantic salinity at 5 meters depths :
     fill/line/key/level=0.2d/z=5                salinity
    The line option means that one superimposes black contour lines to the contour fill
    with a colour key (key) and with salinity intervals of 0.2 psu (lev=0.2d) at 5m depth
 4) To make a plot over the Caribbean area at 5m depth, select the corresponding sub
     set region/z=5/i=1:400/j=1:200
     fill/line/key/level=0.2d salinity
     And one concludes by overlaying a map :
     go land

  5) Let‟s now make the 2D plot of a MERCATOR surface temperature in the Caribbean on
     September 18 2003 (when the Isabel Hurricane was running by), with a contour interval
     every 0.5 degree Celsius.
cancel view
use “”
show data
set region/z=5/i=1:400/j=1:200/l=110
fill/line/key/level=0.5d/lev=(22.5,31,0.5)                temperature
go land

let‟s plot the r.m.s (i.e. the square root of the variance) of the SST field during one week :
shade/lev=0.2d             temperature[z=5, l=1:7@var]^0.5

To save as a gif file, be sure graphic window is not overlapped by any other window, and

Or use the “import myimage.gif” command on a Linux window, and then capture the graphic
window with the mouse.
To quit ferret, type:
quit or exit

Ferret used to access OPeNDAP servers:

use “”
show data

use “”
show data

use “”
show data

MFS OPeNDAP (only Mediterranean basin)
use “”
show data

Using the LAS and OPeNDAP technology to examine the ocean response to
the Isabel hurricane.

The MERSEA project gathers several operational systems for the North Atlantic and is using
the OPeNDAP and LAS technologies. You will use the outputs gathered by the MERSEA
project to examine the ocean response to the Isabel hurricane which traveled from north of the
Bahamas to North Carolina (US) in September 2003.

Description of the MERSEA Inter-comparison project
The Mersea-strand1(2003-2004) and Mersea IP (2004-2008) European project conduct a real
time inter-comparison of the 5 existing state-of–the art operational forecast systems for the
North Atlantic and Mediterranean basins, gathering the following teams (in alphabetical
order): FOAM from UK, HYCOM-US from USA, MERCATOR from France, MFS from
Italy and TOPAZ from Norway. The 5 systems are inter-compared over a one year period
(June 1st, 2003 – June 1st, 2004). Characteristics of the 5 forecasting systems are summarized

           in table 1 below.

MERCATOR    OPA 8.1                      horiz. 1/15° (5-7km)         Daily ECMWF forcing        OI SOFA                                   ATL
            Z coord. + Rigid Lid         verti. 43 levels             Relaxation to Reynolds     SLA SSALTO-DUACS along track
                                                                      SST and Reynaud SSS        once              a       week            MED
            Simple    thermo.      ice   Atlantic +Mediterranean                                 (Jason1,ERS2/Envisat,GFO)
            model                        basin from 10 to 70°N.       Monthly river runoff
                                                                                                 MSSH from Rio et al.(2003) in the
            SPIN UP 15days               Relaxation to Medatlas       Data        assimilation   Atlantic and blend of previous runs in
                                         (T,S) in Gulf of Cadiz       stopped in coastal area    Mediterranean basin.
            TKE                          below 500m                   at depth 500m
TOPAZ NO    HYCOM 1.0                    horiz. 20 to 30km            6 hourly ECMWF             EnKF                                      ATL
                                                                      forcing      (Bulk
            Hybrid coord + Free          verti. 22 hybrid layers      formulae momentum          SLA SSALTO-DUACS Maps once a
            surface                                                   and heat)                  week (Jason1,ERS2/Envisat,GFO)
                                         Artic+Attlantic till 40°S.
            dyn./thermodynanic     sea                                Precipitation Clim, +      SST from CLS AVHRR data once a
            ice                          No Mediterranean basin.                                 week
                                                                      Relaxation to Levitus
            SPIN UP 20years              Closed          boundary     SSS(60days)                Maps of ice concentration
                                         without relaxation.
            KPP mixing                                                No river runoff            MSSH from previous OCCAM run
                                                                      Data        assimilation
                                                                      stopped in coastal area
                                                                      at depth 300m
FOAM UK     Brian-Cox      HADLEY        horiz. 1/9° (12km)           6 Hourly NWP-UK-           OI Cooper&Haines                          ATL
            CENTRE                                                    Met-Office forcing
                                         verti. 20 levels                                        SLA SSALTO-DUACS along track
            Z coord. + Rigid Lid                                      Relaxation to Levitus      once              a       week            MED
                                         Atlantic +Mediterranean      SST and SSS.               (Jason1,ERS2/Envisat,GFO)
            dyn./thermodynamic sea       basin from 10 to 70°N.
            ice                                                       No river runoff            SST 2.5° gridded (ARGO) Once a day.
                                         No Mediterranean Water
            SPIN UP 5months              relaxation                   Data        assimilation   (T+S) vertical profiles at all depths
                                                                      stopped in coastal area
            Kraus-Turner                                              at depth 300m              Maps of ice concentration
                                                                                                 MSSH from previous run with Singh and
                                                                                                 Kelly climatology in Gulf Stream area.
MFS IT      MOM 1.0                      horiz. 1/8°                  6 Hourly ECMWF             OI SOFA                                   MED
                                                                      forcing      (Bulk
            Z coord. + Rigid Lid         verti. 31 levels             formulae momentum          SLA SSALTO-DUACS along track
                                                                      and heat)                  once              a       week
            no ice model                 Mediterranean        basin                              (Jason1,ERS2/Envisat,GFO)
                                         only                         Relaxation to satellite
            SPIN UP:7years                                                                       SST and T vertical profiles along track
                                         Transport        through     night time SST and SSS
            constant vertical mixing                                  climatologies              once a week
                                         Gibraltar parameterized
            + vertical adjustment                                                                MSSH from previous run with 1993-99
                                                                      No river runoff
                                                                      Data        assimilation
                                                                      stopped in coastal area
                                                                      at depth 1000m
HYCOM US    HYCOM 2.1                    horiz 1/12° (6.5km)          3 hourly NOGAPS            OI Cooper&Haines                          ATL
                                                                      forcing          (Bulk
            Hybrid coord + Free          verti. 26 hybrid layers      formulae for heat)         SLA MODAS Maps once a week
            surface                                                                              (Jason1,ERS2/Envisat,GFO)
                                         Atl+Med from 28°S to         SSS= 50%(E-P) +50%
            no ice model                 70°N, 98°W to 36°W.          relaxation to Levitus      MSSH from previous 1/12° MICOM run
                                                                      SSS                        with perpetual ECMWF forcing
            SPIN UP 15years              Entrainment
                                         parameterization of Med      Relaxation to MODAS
            KPP mixing                   Water Outflow                SST analysis.
                                                                      Monthly river runoff
                                                                      Data        assimilation
                                                                      stopped in coastal area
                                                                      at depth 300m
           Table    1:      overview    of     the      various     ocean       forecast   systems
           Foam,   Mercator,     MFS    are   covering     the   Mediterranean      basin  (MED),
           Foam, Mercator, Hycom-us and Topaz are covering the North Atlantic ocean (ATL).

Description of Internal Metrics
The Mersea final aim is to build a European GMES (Global Monitoring for Environment and
Security) operational system in 2008. All systems provide analysis and near real-time
prediction of the three-dimensional ocean. They have built a common methodology, defining
a common grid on to interpolate their outputs and internal metrics to allow a fair inter-
comparison from one system to another. The common geographical domain extends from
10°N to 68°N for the North Atlantic and covers the whole Mediterranean Sea excluding the
Black Sea from 6°W eastward. Internal metrics, i.e. a mathematical definition of chosen
diagnostics, have been defined for the North Atlantic Ocean and the Mediterranean Sea and
aim at testing the consistency, quality and performance of each system. The inter-comparison
exercise runs over a one year period from June 1st, 2003 to June 1st, 2004. Metrics were
defined for the North Atlantic and the Mediterranean basin. There are sorted into 4 classes
(Class 1, Class 2, Class 3 and Class 4). In this exercise, you will use Class1 (through LAS
and OPeNDAP) and Class2 (through OPeNDAP only) diagnostics.

Class 1 diagnostics gather 2-D and 3-D fields interpolated on the MERSEA grid. Two
dimensions fields are the zonal and meridional wind stress (Pascal), the total net heat flux
including relaxation term (Watt/m2), the freshwater flux including relaxation term (kg/m2/s),
the Barotropic Streamfunction (henceforth BSF) (Sverdrup=1e6m3/s), the Mixed Layer Depth
(henceforth MLD) (m). Two kinds of mixed layer depth diagnostics are provided in the
Atlantic basin: MLD(θ) with a 1°C criteria and MLD(ρ) with a 0.05kg/m3 sigma criteria. In
the Mediterranean Sea, only one MLD(ρ) with a 0.011kg/m3 sigma criteria is provided. Other
2-D Class1 diagnostics are the Sea Level Anomaly (henceforth SLA) (m), the Mean Sea
Surface Height (henceforth MSSH) (m) used as a reference during the assimilation procedure.
Three dimensions fields are potential temperature (°C), Salinity (psu), zonal and meridional
velocity fields (m/s), the climatological potential temperature (°C) and salinity (psu) fields
from Reynaud et al. (1998) and Levitus (2001) climatologies in the Atlantic and from
Medatlas (2002) climatology in the Mediterranean basin.

Class 2 diagnostics are potential temperature (°C), salinity (psu), zonal and meridional
velocity fields (m/s) interpolated on high 10km resolution vertical sections and at moorings
locations (figure 1). Some of the chosen tracks coincide with oceanographic cruises or ship of
opportunity tracks.

                            Figure 1: Tracks of Class2 sections

The Mersea common environment- OPeNDAP server and Live Access
All the systems interpolate their outputs on the MERSEA grid with a horizontal resolution of
1/8° and a vertical resolution with 8 vertical levels (at 5, 30, 50, 100, 200, 500, 1000 and
2000meters) in the Mediterranean basin and 12 (at 5, 30, 50, 100, 200, 400, 700, 1000, 1500,
2000, 2500 and 3000 meters) in the North Atlantic. Outputs fields with a common Netcdf
format are delivered through each system‟s OPeNDAP server which URL addresses follow: for FOAM
user=foamessc, password=oceanessc for HYCOM-US for MERCATOR
user=projet, password=rotacrem for MFS for TOPAZ

Above the Opendap technology, we have a Live Access Server (LAS) for class 1 outputs

which federates the access to the model outputs regardless of differences of individual
Opendap servers. Mersea LAS URL is the following: (restricted access) user=mersea1, password=dnarts
A user can quickly obtain products such as plots, images, and formatted files, from custom
subsets of variables – any t, z, y, x combination – (with possibility there to customise the
plot). LAS has a comparison mode which allows the user to select any data sets distributed
on Internet via Opendap and link to the LAS, then to compute differences (with automated re-
gridding), overlay them graphically and view them as side-by-side plots.

How to access an OPeNDAP servers with MATLAB, IDL, GraDS,
FERRET clients…..
Opendap client exists for many application packages (e.g. NCBrowse, Excel, Matlab, IDL,
Ferret, GRADS, IDV, Las, Map Server). The data request functions use the http protocol,
sending an enhanced URL to the server. For Matlab, IDL, and some other Clients, you will
have to install a free software to be able to use the Matlab/DODS and IDL/DODS client.
More information on how to install the OPenDAP clients can be found under the following

Here are some examples for:

      Accessing data using your copy of ferret

       use ""

      Accessing              data             using             your              copy             of            matlab

      Accessing data using your copy of IDL

       url = ''

       ce = '?lons'

       stat = GET_DODS(url, data, DEFLATE=0, /WWW_ERRORS, ce=ce)

      Accessing data using your copy of GRADS


       display t2m

      Accessing data using your copy of Unidata's Integrated Data Viewer (IDV)

       Just paste the selected URL into the IDV data loader.


Some websites where you will find example of the use of Matlab, Ferret, IDL client:
Matlab is the Client releasing the best documentation on how to use the Matlab/Dods
Command line client (i.e. the “loaddods” command). You will find further explanation on the
“loaddods” command under:

For other Clients as IDL, Grads and Ferret, search on google with key words as “ferret

Exercice 1
Here are several useful websites in Oceanography than you need to visit. Please spend some
time to understand what you can find on each web site and how you could use the available
     Visit the following LAS server
     1.   MERSEA LAS: (restricted access) user=mersea1, password=dnarts
     2.   ARCTIC LAS :
     3.   AVISO LAS :
     4.   IFREMER LAS

     Here are some more LAS address you don’t need to visit today:
On the following web page:, you will find an non exhaustive list of LAS Servers.
     5.   LAS sites at NOAA/PMEL :
     6.   LAS In-situ data server providing access to NODC's World Ocean Data Base :
     7.   LAS GODAE Modelers:
     8.   LAS for PMEL's Thermal Modeling and Analysis Project: model results, gridded data products, and research-quality reference
          fields                                                                                                                   :

     9.   LAS for Facility for the Analysis and Comparison of Tsunami :
     10. LAS for coastal bathymetry :

   Visit the following OPeNDAP server
   1. for FOAM
        user=foamessc, password=oceanessc
   2. for HYCOM-US
   3. for MERCATOR
        user=projet, password=rotacrem

   4. for MFS
   5. for TOPAZ

   Visit the following web sites
   1. and visit the “remote sensing”, “ocean modelling”, “data search” sections on the left

The IRI/LDEO Climate Data Library which
make several Datasets available are sorted by category: Air-Sea Interface, Atmosphere,
Climate Indices, Cloud Characteristics and Radiation Budget, Fisheries, Forecasts, Historical
Model Simulations, Hydrology, Ice, Oceanography, Topographic and Land Characteristics.
This Library also gives you directly the command line to download data directly to programs
like matlab, Ferret, GRADS, etc. See for example:

Exercice 2
    Use the MERSEA OPeNDAP with ferret as well as the LAS server to examine the ocean
response to the Isabel hurricane, which took place from Wednesday September 10 2003 until
Friday September 19 2003.
Here is a plot of the hurricane trajectory and a typical forecast bulletin provided to the media
by the Miami Hurricane Center:

Ferret used to access OPeNDAP servers:

use “”
show data

use “”
show data

use “”
show data

Mersea LAS URL is the following: (restricted access) user=mersea1, password=dnarts

Characterize the ocean response to the passage of the Isabel hurricane in terms of SST,
temperature profiles, temperature cross section, Mixed Layer depths, Wind Stress
components, currents, SSH, wake of the Hurricane in SST plots. Here are a few hints for what
you should look for:
     + SST decreases along the path of Isabel starting Sept. 16. The maximum
       temperature drop is 4 degrees.
     + Mixed layer depth increased on the edges of hurricane and decreased close to the
       path of the hurricane center.
     + Strong vertical advection (“upwelling”) behind the eye of Isabel.
     + Hurricane „smoking signature‟ in the mixed layer currents.
     + Thermocline is depressed along the track (≈50m in 5 days).


To top